Simple setup tests bit-error rate

Traditionally, the reception quality of a digital receiver is expressed in terms of BER (bit-error rate). This figure is the proportion of received bit errors in a given period. Typically, you measure the BER in the lab by applying an RF signal, modulated by a pseudorandom code, to the receiver under test. This Design Idea suggests an alternative method based on the use of a simple square wave. This method may not be superior to the usual technique, but it is simple to implement and gives a reliable result. The simplicity of the method is based on the fact that it requires no complex synchronization. Admittedly, a square wave is not truly representative of the data a receiver encounters in normal use (Figure 1). The square wave to modulate the RF carrier is phase-shifted to allow for the delay in the receiver. An exclusive-OR gate produces a sampling pulse at each bit transition—typically, 10% of the data-bit width. This sample pulse samples the raw data the receiver generates, producing clean data.

The key to understanding this technique is to keep in mind that a string of two successive ones or zeros indicates an error. A D flip-flop implementing a 1-bit delay detects the error.

You can display error pulses on an oscilloscope or count them by using a frequency counter. Figure 2 shows a typical test setup. You modulate the RF generator at the prescribed data rate. Note that a 500-Hz square wave is equivalent to a baud rate of 1 kbps. Both the modulating signal and the received data feed into the BER-test board. You adjust the sampling signal to be near the end of the received-data pulse. In many digital receivers, this arrangement yields a fair approximation to a correlation receiver. Error pulses appear on the oscilloscope. If you wish, for example, to set the RF level for a BER of 1-to-100, you reduce the RF level to the receiver such that, in a 100-msec sweep you see, on average, one error pulse per sweep.

In Figure 3, IC₁ and potentiometer P₁ form the basis of an adjustable phase shifter. R₂ provides hysteresis, and R₁, C₁, and IC₂ form a differentiator that provides a sampling pulse train. The first flip-flop clocked by the sampling pulse makes a hard decision concerning each bit. The next D flip-flop, together with exclusive-OR gate IC_2B detects the occurrence of two successive identical bits. This situation constitutes an error. A final D flip-flop and a transistor ensure that the Error output is clean. The construction of the system follows the circuit diagram in Figure 3. It sets an HP8647 RF signal generator at 868.35 MHz, and a function generator provides OOK (on/off-key) modulation. The receiver under test was a Melexis (www.melexis.com) TH7122 at 868.35 MHz in the OOK-modulation mode. Adjust the RF level to vary the error rate. This design obtains an RF level of –107 dBm for a 1-to-1000 BER and –108 dBm for a BER of 1-to-100, levels consistent with the data sheet. You should take care when you're implementing OOK. Most RF generators provide AM. Thus, you must remove 3 dB from the displayed RF value. You can use this technique for other types of binary modulation, such as FSK (frequency-shift keying), for example.

Check out our Best of Design Ideas section!